We’ve been having some problems with our DSL service at Chateau Steelypips again, which has gotten me thinking about the design of devices that are annoying to use. It occurs to me that you might use a sort of control to indicator ratio as a measure of how irritating a device is to use.
This is prompted by the fact that our home network contains two devices, a DSL modem and a wireless router, each of which has four indicator LED’s on the front. These LED’s have a couple of different states each– when either devices is functioning properly, they’re steady green, but they can also blink, and light up in different colors. Presumably, different failure modes are distinguishable by which lights are lit, and which are blinking, and so on.
Each device also boasts only one control: a power switch. That’s it. I have no real idea what the different blinking lights mean, because it really doesn’t matter. Regardless of how the device has failed, there’s only one action you can take: you can turn it off, and turn it back on, and hope for the best.
This is, of course, ridiculously frustrating from a user standpoint. There’s no point in having the ability to distinguish between 81 different states of the device (four lights, each eith on, off, or blinking) if there’s only one thing you can do in response. It’s not like I can say “Ah, the third light from the left is out, that means that I need to renooberate the frobulator…” and do that. Whichever light is blinking, the only thing I can do is turn the modem off, and turn it back on.
Hence, the idea of the control-to-indicator ratio as a measure of the annoyingness of a device. The ideal value would be close to 1– not necessarily having each indicator adjusted by a single control, but having multiple actions possible for the user to get the indicators into the right alignment would be helpful. The control-to-indicator ratio of 1:4 (C/I = 0.25), for our DSL modem is Not Good.
Of course, being far from one in the other direction isn’t good. A device with four knobs and only one indicator light (C/I = 4) isn’t going to be any fun to use, either– you’d have a great deal of freedom to make adjustments, but you’d be flying blind most of the time. So maybe we need some more complicated formula, to measure the deviation from one in either direction…